Research · Evaluation · AI-assisted analysis
I design studies, evaluate programs, and build AI workflows that turn thousands of survey responses, interviews, and job postings into findings your team can use. Most of my clients are workforce programs, colleges, and universities with more data than they have time to analyze.
Labor market and skills analysis, curriculum alignment, and stakeholder studies for workforce and STEM education programs
Program evaluation, curriculum and credential design, and facilitation
Validated LLM workflows for any large text dataset: survey comments, interviews, job postings
Published, peer-reviewed research and current client work
Current engagement · 2026–27
Literature, stakeholder interviews, and case study analyses to inform the design of a national apprenticeship program from both practical and pedagogical perspectives
Journal of Engineering Education · 2024 · Lead author
Analyzed 26,103 engineering job ads against the O*NET skills taxonomy to compare what employers ask for with what programs teach
Read the paper →Humanities & Social Sciences Communications · 2026 · Co-author
Co-developed GATOS, an open-source LLM workflow for building qualitative codebooks, and tested it against traditional thematic analysis
Read the paper →I’ve always been most interested in method: how you know what you think you know. In my PhD, that meant fabricating suspended micro-devices and figuring out how to transfer nanomaterials while maximizing area and minimizing thickness.
Since then, “methods” has meant interviews, surveys, and AI workflows for analyzing qualitative and mixed methods data at scale, mostly on how engineers are trained and hired. The questions and methods might have changed, but my core belief has not: outcomes are only as good as the methods you use to obtain them.
Gabriella Coloyan Fleming, PhD
Founder & Principal
I’m happy to talk it through, even if you’re not sure yet what kind of study it needs